Brightest cluster and intracluster light - Observations Magda Arnaboldi European Southern Observatory (Garching) O. Gerhard, L. Coccato, G. Ventimiglia, S. Okamura

P. Das, M. Doherty, K. Freeman, E. Mc Neil, N. Yasuda, J.A.L. Aguerri, R. Ciardullo, K. Dolag, J.J. Feldmeier, G.H. Jacoby.

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 1 Outline

1. Brightest Cluster Galaxies and Intracluster light in clusters 2. Planetary nebulae as kinematical tracers 3. The Virgo cluster & M87

– The PNs’ VLOS and the projected phase space distribution – Dynamical status of the Virgo core – Large scale distribution of the ICL 4. The Hydra cluster & NGC 3311 – Observations : long-slit spectroscopy, MSIS, surface photometry – Debris from disrupted galaxies by the cluster potential 5. The core & the NGC 4874/NGC 4889 binary merger 6. Conclusions

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 2 1. ICL from deep photometry • Core of the Coma cluster: Photographic photometry Thuan & Kormendy 1977

• Abell 4010 (left, z=0.096) Abell 3888 (center, z=0.151) Deep CCD photometry Krick & Bernstein 2007 Abell 1914 (right) Feldmeier+2004 See also Melnick+’77, Bernstein ’95, Feldmeier+’04, Gonzalez+’05, Mihos+’05, Krick+’06

See contributions from J. Krick, J. Melnick, C. Rudick, S. Okamura M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 3 ICL properties in individual clusters

• ICL surface brightness profile shape varies between clusters - Krick & Bernstein 2007 • Ellipticity generally increases with radius, position angle sometimes has sharp variations - Gonzalez + 2005 • Suggests ICL is dynamically young and separate from BCG

PA (o)

50 0 -50

1 10 100 (kpc/h) 4 M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 ICL is ubiquitous – statistical properties

• ICL has been found in – BCG dominated clusters (e.g., Gonzalez+’05) – in clusters without BCG (e.g., Feldmeier+’04, Struble88)

• Average ICL+BCG and total (incl. galaxies) cluster surface brightness profiles, from stacking 683 (scaled) clusters from SDSS (Zibetti+’05): – ICL (=SB excess w.r.t. R1/4 profile of central BCG) is more centrally concentrated than cluster galaxies – ~11%; BCG ~22% (**) – ICL SB correlated with BCG luminosity and cluster richness, but ICL fraction almost independent

• ICL fraction in individual clusters varies: – ~4-21% (in B, Krick&Bernstein 2007) – ~10-30% (Feldmeier+’04) – ~5% in Virgo, isolated regions – ~50% in Bernstein+’95 field in Coma cluster core Depends on radial range and evolutionary stage of the cluster (**) robust values, measured from profiles

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 5 BCGs & ICL

Dolag+2010 o Disentagle cD from ICL in clusters - use the velocity distribution of . o Simulations of clusters - VD is bimodal. Fit by two Maxwelliansdistribution, one narrower (colder) for the central and a broader one (hotter) for the ICL. o Hotter component is responsible for the “light excess” at large radii. o Stellar particles in hydrodynamical cosmological simulations thus selected turn out to have different spatial distribution & formation history. See also Kapferer et al., 2010, MNRAS, 516, 41.

Kinematics measurement can tag galaxy and ICL stars from their LOSVD at the same spatial location !

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 6 2. Planetary nebulae as kinematical tracers • PNs trace light because the luminosity-specific stellar death rate should be independent of the precise state of the underlying stellar population (Renzini & Buzzoni 1986), see also PN.S results from Coccato+09. • PNs are distance indicators. • The [OIII] line emission at 5007Å is the strongest emission from a PN; it allows the identification & the measurement of its radial velocity ON-[OIII] ON-H OFF-(V+R)

• We obtain PN number density distribution and 2D radial velocity fields in regions where the stellar surface brightness is too faint with respect to the night sky!

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 7 3. The core of the Virgo cluster

Mihos et al. 2005, ApJ, 631, L41 & ESO PR 2009/19a -1 And this is where it all started: 3 PNs at vmean ~ 1400 kms along the LOS to -1 NGC 4406 (vsys = - 240 kms ) Arnaboldi, Freeman, et al. 1996, ApJ, 472, 145

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 8 3. The core of the Virgo cluster

Narrow band imaging plus spectroscopy

ICPN catalogs from Arnaboldi+02&03, Feldmeier+03&04

6 VLT-Flames fields Core, FCJ, SUB from Arnaboldi+04 F7_1,F7_2, F4 from Doherty et al. 2009, A&A, 502, 771

52 PNs vlos total

M87 halo according to Kormendy+09 photometry

**For optical and Spitzer IRAC surface photometry in Virgo see also oral presentations of C. Rudick and J. Krick .

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 9 See also Freeman et al. 2000, ASP 197, 389, and Arnaboldi et al. 2003, AJ, 125, 514 for spectroscopic confirmation of ICPNs.

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 10 PNs’ LOSVD and projected phase space

M87 halo

• does not rotate. From GCs’ vrot -1 expects vave = 1150 kms for M87 halo PNs. M87 ICL •σ(r) decreases. From Montecarlo simulations of a Gaussian with σ=247 kms-1, the combined probability for 2 measurements at > -2σ and 5 within +/- 0.3 σ is less than 1%. ICL

•No ICPN vLOS redder than 1800 kms-1 at R<3600”, while bluer velocities present at all radii. -1 •Broad velocity component at σPN = 247±52 kms @ 60 kpc -1 R>3600”. σPN = 78±25 kms @ 144 kpc

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 11 PNs’ LOSVD and projected phase space A04 ICPN detections Dwarfs spheroidals Red and blue symbols indicate the Doherty+09 PN detections.

PNs with velocities higher than the mean velocity of the Virgo cluster (1100 km s−1) are shown as + ; those with lower velocities are shown as ○.

The PNs associated with the peak at the M87 vsys are spatially segregated . There are no PNs at the M87 velocity at R > 150 kpc, while ICL PNs are scattered across the whole region! M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 12 Dynamical status of the Virgo core

• The LOSVD of dEs+dS0s is flat and broad with a peak at M87 (1300 -1 kms ) with tails at vLOS<0 (Binggeli+87); asymmetry in the ICPN LOSVD in the Virgo core center. • Observed properties of the ICPN LOSVD are predicted in merging simulations of two sub-clusters with unequal mass along the LOS shortly before ~ 109 yrs the sub-clusters merge (Schindler & Böhringer 1999) . • Recent and on-going assembly of Virgo core: M87 falling backwards from the front and M86 forwards from the back (Binggeli+93). • The light in the M86 sub-group is tidally stripped by the more massive M87 component, while these two merge along the LOS - • We do not see a diffuse light with a broad velocity component redwards

of M87 vsys because it has not yet been formed.

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 13 Large scale distribution of ICL in the Virgo cluster

• RGB counts • PNs Arnaboldi+02,03,Aguerri+’05 • PNs Castro-Rodriguez+’09  PNs Feldmeier’04 M49

Radial Surface brightness from Virgo galaxies Bingelli+’87 ( - total, -.- giants, -…- dwarfs )

PN surveyed fields in the Virgo cluster (Castro-Rodriguez+09): PN number density is translated in B band. Surface Brightness (SB) and the large scale SB profile is plotted as function of the distance from M87 out to 4 degree = 1Mpc.

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 14 Large scale distribution of ICL in the Virgo cluster

Surface number density maps for the globular clusters in comparison with the spatial distribution of galaxies and x-ray emission in the Virgo cluster region. DDEC and DRA are distances from M87 along declination and right ascension, respectively (Lee et al. 2010, Science 328, 334). M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 15 Large scale distribution of ICL in the Virgo cluster

ICL distribution is clumpy & associated with the high density peaks in the Virgo cluster. There is not a smooth large scale light distribution out to the Virial radius of the Virgo cluster detected so far.

..A cautionary note for extrapolation of core ICL profile out to Virial radius for estimates of the baryonic fractions...

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 16 The next step beyond Virgo and Fornax ….

The brightest PNe in the Coma cluster at  100 Mpc distance have fluxes of 2.2  10-18 erg/s/cm2 (~2 photons/min on 8m tel.). They cannot be detected using a narrow band filter because the noise in the 30 – 40 Å sky is of VLT + the same order of the signal we want to detect. FORS2 Subaru + FOCAS

Ventimiglia et al. 2011, A&A, 528,24

Multi Slit Imaging Spectroscopy Technique (MSIS) Blind search: it combines a mask of parallel multiple slits with a narrow band filter, centred on redshifted [OIII] 5007 Å line at Coma mean systemic velocity, to obtain spectra of all PNe that lie behind the slits. The sky noise at the emission line of the PN come from few Å only, depending on slit width and spectral Gerhard et al. 2005, ApJ, 621, L93 resolution…

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 17 5. The Hydra I cluster & NGC 3311

Observations of the Hydra I core • Hydra I : example of a relaxed cluster • LSS - Long slit data acquired with FORS2 at VLT – GRISM-1400V+18 (~0.75Å/pixel) (*) • MSIS - Data acquired with FORS2 at VLT (*) – GRISM-600B (~ 1.5 Å pixel-1 ) plus two narrow band filters centered at λ = 5045 Å and λ = 5095 Å, with a FWHM of 60 Å – One field centered on NGC 3311 – FoV of 6.8 x 6.8 arcmin2, equivalent to 100 x 100 kpc2 . – 56 PNs detected & 26 background galaxies • Surface photometry - V band imaging at the ESO/MPI 2.2m telescope and FORS1 archive data.

(*) see Coccato and Ventimiglia’s posters M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 18 5.1 Long slit spectroscopy

• NGC 3311 - Central galaxy velocity dispersion 150 kms-1 • In 15”-45” (4-12 kpc) transition to 450 kms-160% of dispersion of cluster galaxies (*) • Outer halo unusually dynamically hot for ETGs

LS-spectra from FORS2 (082.A-0255(A) ,4655 – 5965 Å) and GMOS archive (3675-6266Å) (*) see independent measurements from Richtler+11 (arXiv1103.2053) (Ventimiglia et al., 2010, A&A, 520, L9) M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 19 5.1 Long slit spectroscopy

The dynamically hot stellar halo around NGC 3311: a cluster dominated small galaxy Properties of the early-type halos: • * Halos mapped using PNs (Coccato+09) • ◊ Coma BCG galaxies NGC 4889 & NGC 4874 (LS spectra Coccato+10) • ♦●■ measurements of σ for NGC 3311 at the galaxy center, 15” (3.7 kpc) and 47” (12 kpc).

Intracluster stellar halo dominates completely for R>12 kpc

• Halo was accreted – origin not co-eval with inner (R< 3.7 kpc) regions

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 20 5.2 MSIS observations of the Hydra I core

GCs vLOS

Hydra PNs LOSVD m5007 vs vlos diagram for the Hydra PNs There are two additional velocity Comparison of the observed PNS subcomponents in the observed LOSVD with the simulated Hydra PN LOSVD: cluster PN LOSVD ( a Gaussian a blue (1800 kms-1) and a red -1 VD centered at 3709 kms , with (5000 kms-1 ) secondary peaks. σ = 500 kms-1 ). GCs LOSVD not Gaussian PNLF @ D=50 Mpc convolved (Richtler+11) with similar with slit losses and filter secondary peaks! transmission. (Ventimiglia et al., 2011, A&A, 528, 24) M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 21 5.3 Surface photometry

WeRed now PNs lookdistribution for the light peaks substructures in the NE quadrant corresponding of the NGC to the 3311 velocity halo…. subcomponents

PA 32

2D GALFIT best Sersic fit with index n=6.3

•Excess of light in the NE quadrant •Total luminosity 9 8 Lexcess = 3.2 ·10 L (±5.2 ·10 )L •2D surface brightness profile for

NGC 3311 : Re = 11 kpc (from light- growth curve) •Total luminosity in the dwarfs 9 LDWs= 1.1 ·10 L -ESO/MPI 2.2 m telescope WFI data – FORS1 V band imaging (ESO archive) - (Arnaboldi+2011 A&A submitted.) M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 25 5.4 Light Excess in the Halo of NGC 3311

-1 • DWs’ Vlos = 4479 - 5735 kms Residual image •Concentrated in the NE V band – 2D Sersic n=6.3 quadr. & at the same location of the red peak PNs. •No other DWs within (100 kpc)2 of NGC 3311 •Light excess is 12% of light in slit region.

II I HCC26

Spectra of light excess obtained from slit regions I,II as Sex= S1D – 0.9 * SNGC 3311 , where SNGC 3311 is halo SE template

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 26 5.4 Light Excess in the halo of NGC 3311 – cont.

-1 -1 In spectra I and II , vLOS of the light excess is at 5032 kms (±38 kms )!

I II

-1 •The velocity of HCC26 (vsys = 4225 kms ) and those of dwarf galaxies in the same regions as the PNs are consistent with the red subcomponent at ~ 5000 kms-1 in the PNs LOSVD. Light excess at the same location and LOS velocity. •Blue peak traces the NGC 3311 halo.

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 24 5.5 The Hydra I cluster – Results

• The stellar halo of NGC 3311 becomes hotter at large radii - accreted origin. • Presence of discrete velocity subcomponents in PNs LOSVD - substructures not phase mixed yet (*). • The PNs in the red peak are associated in velocity and space -1 with a group of dwarf galaxies at vmean = 5000 kms • Presence of a light excess at the same location & vLOS of dwarfs galaxies and red peak PNs in the halo of NGC 3311. 9 • Lexcess= 2.3 ·10 L , a factor 2 times luminosity in dwarfs • Results consistent with the light excess tracing debris by disrupted dwarfs galaxies as they fall and interact with the Hydra I core gravitational potential (**).

(*) Poster from G. Ventimiglia. (**) see L. Coccato’s poster on stellar populations in BCGs

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 25 6. Diffuse light in the Coma cluster Coma is one of the best studied nearby clusters; of these, it is the richest and the most compact one. It provides a laboratory for studying the effects of a dense environment on galaxy evolution.

Presence of diffuse light is established from SB measurements: •extended halo around NGC 4874 and NGC 4889 (Thuan & Kormendy 1977; -1.3 Bernstein+95). µv ~ r for r > 40” from center of NGC 4874 •This extended halo is not uniform: there is presence of plumes and tidal tails (Gregg & West 1998, Adami+05a) •Outside this halo, giant arcs and debris are also detected (Trentham & Mobasher 1998, Gregg & West 1998) 10 • (*) Recent surf. phot data – ICL LB = 7.6  10 L◉ contrib. by Adami, Durret, Okamura - & GCs – E. Peng Diameter ~200 kpc Thuan & Kormendy 1977 M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 26 PNs LOSVD in the Coma field

• 37 PNs detected •Their LOSVD is not a single Gaussian •Multi peaked – secondary bluer peak associated with nearby substructure G7 •Main peak of the PN LOSVD is at 6450 kms-1

G7 sub-group VLOS < 5500 km/s Adami+05b, A&A 443, 17

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 27 PNs LOSVD in the Coma field

NGC 4889 NGC 4874 • 37 PNs detected •Their LOSVD is not a single Gaussian •Multi peaked – secondary bluer peak associated with nearby substructure G7 •Main peak of the PN LOSVD is at 6450 kms-1 • Coma core is not virialized

NGC 4889

NGC 4874

-1 NGC 4874 0 = 300 kms -1 NGC 4889 0 = 400 kms

We measure the vLOS of stars in the diffuse stellar halo in a region near to NGC 4874, but stars’ vLOS histogram is peaked at -700 kms-1 from its systemic velocity - Halo velocities flipped!

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 31 Results - the sub-cluster merger in the Coma Cluster

Summary The sub-clumps in the galaxy distribution centred on the two supergiant galaxies (Fitchett & Webster 1987, Biviano et al. 1996) have the average velocity of the other BCG (Colless & Dunn 1996) => Halo velocities flipped. Elongation of ICL and the MSIS PN kinematic measurements make a strong case that this ICL is in highly dynamically evolving state. Early binary model (Valtonen & Byrd 1979). Head on collision of cDs (Korchagin,Tsuchiya &Miyama 2001), Implications In these nearly head on collisions, the outer envelopes of both structures merge in a slow oscillatory fashion along an essentially unchanging orbital direction. By contrast, the interaction of the cores is faster stronger and may involve a significant impact parameter.

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 29 Results - the sub-cluster merger in the Coma Cluster

With the near head-on ICPN and galaxy geometry of the collision, velocities in it is likely that a pre- surveyed field are existing cooling core in concentrated the cluster would have been heated by the around NGC 4889 collision. This is velocity consistent with the absence of a cooling core in the Coma custer now

Gerhard et al. 2007, A&A,468, 815 Residual X-ray emission and ΔT=300 Myr temperature map indicate arc shock following collision

A similar merger is currently ongoing more nearly along the LOS in cluster CL0958+4702 (z=0.39, Rines, Finn & Vikhlinin 2007). Elongated plume of old stars from BCG seen in Spitzer IRAC images.

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 30 Simulated BCG Merger

Blue galaxy is at (0,0) with “green” velocities Halo velocities flipped z ~ 0.034 |V1 – V2| ~ 1600 km / sec

|V1 – V2|SKY ~ 1200 km/sec D ~ 215 kpc

DSKY ~ 145 kpc

Cluster simulation from K.Dolag, analysis from L. Coccato (2011, in prep.)

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 31 7. Conclusions

• Diffuse light in cluster cores ubiquitous – deep photometry and kinematics from PNs show it is made up of genuine ICL and extended, luminous halos of BCGs. On average, ICL contains ~10% of cluster stellar light, BCG halos further ~10%; substantial variations depending on cluster evolutionary state. • Kinematics from PNs show that ICL is un-relaxed with discrete velocity components (Virgo, Coma, Hydra) over cluster  scale. BCG halos are colder; constant (NGC 4889), decreasing (M87, NGC 1399), or increasing (Hydra, collapsed ICL) ’s. Together with morphology: dynamically distinct components. • Formation of ICL and outer halos, how and when? Evidence for BCG mergers (Coma+) and tidal disruption/accretion (Hydra) in cluster cores – build-up of BCG halos and nearby ICL continues. Formation of ICL and outer halos on- going and long-lasting process.

• Stellar population characteristics of the halos and the ICL? – more work needed lots to learn: see contributions by L. Coccato, O. Gerhard, I. Loubser

M. Arnaboldi – BCGs and ICL: observations ESO, FVC+, 27 June - 01 July 2011 32